|Members log in here with your user name and password to access the your admin page and other special features.
|| SportsShooter.com: News Item: Posted 2008-08-18
An introduction to the strip camera, how Tom Dahlin made his, and how you can too.
By Tom Dahlin
Photo by Tom Dahlin
This image of Johan Santana was used by SI for Kids on the December 2007 issue's table of contents page. Dahlin made it using the modified Nikkormat film camera described in this article.
George Silk first used one to produce an innovative and abstract photo essay for Life magazine of the 1960 US Olympic tryouts. He used it again for another essay on Halloween which resulted in the October 31, 1960 Life magazine cover. John Zimmerman quickly embraced the technology and applied it to make a portrait of Pete Rose. Neil Leifer made regular use of it in the 1970's photographing athletes such as Billy Kidd and Gaylord Perry. Bill Frakes made a stunning image of Marion Jones winning the 100m in the Sydney Olympics using one. I am referring to the strip (sometimes called slit) camera, a rather unique device that produces a characteristic streaked background, and records a time history of an event, rather then a snapshot of an instant. In this note, I am going to explain what a strip camera is, how it works, give a little history, and explain how I made mine and how you can make one for yourself.
Strip Camera Basics
A film based strip camera (see sidebar on digital versions) makes use of a thin slit mask inserted in between the lens and film plane. This slit, typically about 0.004" to 0.015" wide, is oriented perpendicular to the film’s travel path. Light must pass through the slit in order to reach the film. To record an image, the film is set in motion, and the camera’s shutter is opened for a period of a few seconds, depending on the length of the event to be recorded. Because the film is moving, stationary objects in front of the lens produce streaks. But if a moving object passes in front of the lens with a speed and direction in sync with the film’s motion, a sharp undistorted image is recorded. See figure 2 for a simplified diagram of this.
Sidebar – Digital Strip Cameras
Digital versions of the strip camera exist. One is the ubiquitous FinishLynx photo finish system used at most major track events. Designed for timing purposes, it records digital images with relatively low resolution and dynamic range. As such, the images will not impress most editors. Another way to make a digital strip camera is to employ a line scan machine vision camera with a linear (1 dimensional) sensor. Such cameras are widely used in the process inspection world, but require specialized software to assemble an image. The cameras themselves are expensive and tend to have poor light sensitivity. I'm in the process of fooling around with a few different types of these, but am having trouble finding the time to complete the project. I've gotten a couple of these cameras to record images, but have found the image quality lacking when compared to what my film camera can do, But it’s only a matter of time before these cameras get better and cheaper.
What makes a Strip Image Different
A slit camera operates by moving film past a thin slit located in between the lens and the film. This diagram shows how I connected a motor to the rewind knob of a 35mm camera to move the film.
A number of factors contribute to the uniqueness of a strip camera image. The principal ones are:
• Distortions: The Picasso like distortions of arms and legs of runners are characteristic of strip camera photos. Objects such as swinging arms and pumping legs travel across the slit at different rates then the body torso to which they are attached. This results in elongation or compression.
• Clean backgrounds: One thing I really love about strip camera photos is that there is no such thing as a crappy background. Everything that doesn't move is rendered as a colored streak. Often you can find a very colorful patch of background that compliments your subject, and set up the camera so that the slit lines up with chosen colors.
• Lack of perspective: Consider a strip image of a pack of runners on a track. At first glance, it does not appear to be much different then an image made with a conventional camera, except for the streaked background. But take a closer look at the runners at the front and back of the pack, i.e., on the right and left edges of the print. You cannot see their chest or back. All of the runners look as if they were photographed directly from the side, which indeed they were! When the eye views this kind of a strip photo, it recognizes that something is different, but often cannot pinpoint what the anomaly is.
• It never happened. Read that again and repeat. It never happened. The image a strip camera records is a time history of what crossed in front of the camera’s slit. Unlike most cameras, this is not a snap shot of an instant in time. What is recorded by the strip camera never happened in an instant, as our human perception wants to interpret it. For instance, think of an image of hurdlers in action. A strip camera image shows all of them up in the air, crossing the hurdle. The lead guy is to the left, and his pursuers follow in order. In reality, there was no time in the race when all of the hurdlers were in the air simultaneously, although the photo shows this.
The strip camera has it’s technology roots in panorama photography where the concept of exposing film through a slit on a swinging lens goes back to 1844 when Friedrich von Martens, a German living in Paris developed such a camera. The Kodak Cirkut panoramic cameras that followed later in the early 1900’s used the same principle of operation (ref 1).
Because a fixed sheet of film exposed through a moving slit is just the relative opposite of a fixed slit with moving film, the above cameras can be considered the great grand daddies of the strip camera family.
In the 1930's horse race photo finishes were recorded using conventional high-speed motion picture cameras. Even at high frame rates (65 fps), these cameras often missed the critical moment. To solve this problem, shutter-less strip cameras were developed. In a 1939 Phototechniques article (ref 2), Ralph Powers described this evolution and discussed the design of his strip camera. Around the same time, Lorenzo del Riccio, head of Paramount Pictures technical laboratories independently developed his own version, which was also intended to be used for photo-finish applications (ref 3). By the late 1930’s, strip cameras became the industry standard for photo-finish recording.
George Silk was a Life magazine photographer noted for his creative and unique style. He was perhaps the first photojournalist to make use of a strip camera, certainly the first that I could find.
In an interview by John Loengard (ref 4), Silk recalls first thinking of the use of a photo finish camera while doing a story on quarter horses in California. For this story he submitted a photo taken by the tracks photo finish camera, from which he was given the negatives. According to Silk, the photo was used but no specific issue is cited. Later the next year, he arranged to borrow a spare photo-finish camera from the Kentucky Derby, which he took home to experiment with, using his kids as test subjects. Recognizing the uniqueness of the images, he asked legendary New York camera repairman Marty Forsher to make a portable version of the camera. Forsher produced a camera powered by a spring wound gramophone motor that used 100 foot rolls of 35mm film. Silk used this camera to cover the 1960 Olympic trials, producing an 8-page photo-essay for Life in the July 18, 1960 issue. He also used the camera to produce the October 31, 1960 Halloween Life cover. Figures 4 and 5 are reproductions of these.
This is the opening two pages of George Silk’s photo essay on the 1960 Olympic Tryouts that appeared in the July 18, 1960 issue of Life magazine. Silk used a strip camera to record unique images.
John Zimmerman is another Time-Life photographer known for his technical wizardry. He immediately saw the creative potential of the strip camera and borrowed (ref 5) a camera from Silk to photograph baseball's 'Mr. Hustle', Pete Rose. Zimmerman went on to use a close cousin of the strip camera, the slit-scan camera, to produce the famous Nate Archibald and Dr. J basketball photos used in Sports Illustrated.
In the 1970’s, Neil Leifer of Sports Illustrated used a strip camera on several occasions to photograph subjects such as Gaylord Perry, Billy Kidd, Indy Grand Prix action, and several other sports. His book Sports!, (ref 6), features four of these images on the dust jacket, photographed between 1969 and 1974. According to the technical notes in the book, he used a Nikon F with a 250 exposure back, converted to a strip camera by Al Schneider of the Life photo lab. In an interview with Larry Berman (ref 7), Leifer discusses the strip camera, referring to it as "a Rube Goldberg made by Al Schneider".
Bill Frakes and David Callow made perhaps the most iconic contemporary strip camera image with a shot of Marion Jones’ 100m finish in the 2000 Sydney Summer Olympics. According to an interview by John McDermott posted on the Canon Web site (ref 8), the camera used was a Hasselblad modified by Callow who served as Frake’s assistant for the event.
More recently, fellow SportsShooter.com member Larry Graves has gotten some good press in Wired Magazine in an article featuring his NASCAR strip camera images (ref 9). Like Frakes and Callow, Graves uses a modified Hasselblad.
Construction of a strip camera
Photo by Tom Dahlin
High School Track. Although the bodies of the subjects are in sync with the film speed, their arms and legs are not. This results in interesting distortions. Note the clean and uncluttered background as well.
You might be thinking 'where can I buy a strip camera'? The answer is that you can't. At least I don't know of any sources. All of the cameras discussed above were made by modifying standard film cameras. These modifications can be very easy or very difficult, depending on the results you want to achieve. If you are after artsy distortions and don’t mind some vertical banding, then a few hours of work and an old film SLR should be all it takes. If you want consistent results and a very even, band free image, then you have quite a bit more work in front of you, likely involving machine shop practice, motor speed control electronics, and laser cutting.
By far the best place to start is by reading the works of Andrew Davidhazy, a professor at RIT with a passion for strip cameras and a prolific writer. In his publication "Basics of Strip Photography" (ref 10), Andrew describes a simple method of using a standard 35mm film camera and converting it to a strip camera. It was this piece that inspired me to try my hand. Do a Google search on Andrew and you’ll find several other publications on strip cameras such as in references 11-13.
The camera I used for the images in this article is shown in figure 6. It's a modified a 35mm Nikkormat, one that I’ve had since my high school days.
I used the technique that Davidhazy developed where the entire roll of film is first advanced to the camera’s take up reel by firing off 24 or 36 blank exposures with the lens cap on. After doing this, the rewind release is pressed and the shutter cocked. To take a strip photo, the shutter is fired on bulb and the film is rewound with the shutter open by turning the rewind knob while holding in the rewind release. I made the slit using a laser cut stainless steel shim and taped it right between the film guide rails under the film. I also employed a microprocessor to control a stepper motor to turn the rewind knob.
The use of a stepper motor is my little twist on strip camera design. I chose to use it because it allows precise and repeatable control of motor speed, is easily interfaced to a microprocessor, and allows me to easily keep track of the number of turns the motor has turned, thus the length of film used. It also permitted me to implement a crude bar code marking at the end of the image, thus allowing me to write the film transport speed to the film for post event results analysis. This is especially valuable when you are experimenting with different film transport speeds. All I needed to do is develop the film and find the best looking results and then read off the transport speed used from the negative.
Film, Processing, and Photoshop
Photo by Tom Dahlin
This is my modified Nikkormat used to take the images in this article. Dahlin used a stepper motor to turn the rewind knob. The motor is controlled by an external control box containing a microprocessor.
I generally shoot Fujicolor 400 or 800 speed color negative film. I use this for the exposure latitude it provides. I have it processed at the local Sam's club for $1.56 per roll. I specify that I want processing only, no prints, and not to cut the film. Following this, I eyeball the processed rolls for good images and then scan them using an Epson 4990 flatbed scanner. Since the images are typically about 3 to 9 inches long, the use of a flatbed scanner allows me to scan an entire image rather then have to piece it together from single frame scans.
Once scanned and into the computer, I apply the usual levels, contrast, saturation and sharpening steps in Photoshop. For some images, use of a slight posterization effect provides a nice boost in contrast and image snap.
Banding is a phenomenon that results from uneven film transport speed. If the film stutters a little while moving past the slit, it will take a little more or a little less time to cross the slit, resulting in exposure differences which show up as vertical bands.
To eliminate or minimize banding, I've developed a technique where I make a layer that contains only the banding modulation and then use an inverse of this layer to subtract the brightness variations from the original. To make the layer containing the banding, I duplicate the main image onto a new layer, and then crop a portion of the image that runs the entire length, but only contains a very limited vertical portion where there should be no variation of the background, i.e., the subject is not in this portion of the image. This image section is usually found on the top or bottom of the image, and theoretically should be an even, constant tone.
I then collapse this cropped portion of the image to a single line of pixels using the transform function, and then expand it to cover the full height of the image. The collapsing forces an averaging of the vertical pixels, and the expansion spreads the averaged values over the entire vertical dimension.
I convert this layer to a gray scale and invert it. The result is a layer containing bands of gray that are opposite (inverted) those in the original image. I then blend this layer with the original using the 'multiply' mode of blending. By playing around with the sliders, I can greatly reduce or eliminate the banding.
Strip cameras have been around for a while, but have not been used very much lately. Perhaps it is due to overuse of the technique in the '60's and 70's, or perhaps it is due to the decline of the film industry and the skilled repair craftsmen who have the knowledge and tools to make the cameras. I do believe that in the next few years that you'll see digital versions of strip cameras emerge with quality exceeding film. I’m sure we’ll see a few of these images from Beijing.
Resources and references
1. Life Library of Photography – The Camera, Time-Life Books, New York, 1970. (Pages 150-151 describe a swing lens camera developed for panoramic photography in 1844 by Friedrich von Martens. The Kodak Cirkut cameras that followed later in the early 1900’s used the same principle of operation.)
2. Photography at the Race Track, Ralph Powers, Electronic Control Corporation, Photo Techniques, June 1939. (The article provides a historical timeline for the introduction of strip cameras to racetrack photo finish use. Interestingly, there is no mention of the work of del Riccio, a possible rival.)
3. US Patent 2320350, Camera, Lorenzo del Riccio, applied for March 20, 1939, patented June 1, 1943. (Patent describes a photo-finish camera employing a slit.)
4. Life photographers: What They Saw – John Loengard, Bullfinch Press, Little, Brown and Co., Boston, 1998 (Silk Interview on pages 175-193 provides history and timeline of Silk’s strip camera use.)
5. Photographing Sports: Capturing the Excitement of People in Action.
John Zimmerman, Mark Kaufman and Neil Leifer. Masters of Contemporary Photography Series, Alskog Book ISBN 0-690-00786-8, 1975. (Page 61 credits Silk as the provider of Zimmerman’s slit camera for the Pete Rose portrait)
6. Sports!, Neil Leifer, George Plimpton, Harry Abrams, NY, 1978. (Technical notes in book describe strip cameras used.)
7. An Interview with Neil Leifer, January 8th, 2002, Larry Berman and Chris Maher, http://www.bermangraphics.com/press/leifer.htm
Photo by Tom Dahlin
This is the motor speed control board. It contains a microprocessor and stepper motor drive components as well as an LCD display and some pushbuttons for operator interaction.
8. Canon Professional Network, Behind the Lens Series. Bill Frakes: Olympian Ideal. http://cpn.canon-europe.com/content/btl/frakes.do
9. Wired Magazine Issue 15.07 – Hacking NASCAR: The Ultimate High Speed Photography Kit. Erin Biba, http://www.wired.com/gadgets/digitalcameras/magazine/15-07/st_nascar
10. Basics of Strip Photography, Andrew Davidhazy, Rochester Institute of Technology. (This should be your starting point for the Davidhazy papers. It begins with an overview of the various types of strip cameras and then proceeds to describe the construction of a simple strip camera using a 35mm camera.)
11. How I Broke the Reality Barrier, Andrew Davidhazy, Popular Photography, October 1970.
12. Strip Photography – How it can be achieved with a 35mm camera, Andrew Davidhazy, The Photographic Journal, Volume 124 No. 1, January 1984.
13. Stretch your Vision, Modern Photography, July 1980.
Other articles on strip cameras
Looking at Life Through a Slit, Milwaukee Journal, June 27, 1971. (An article on the work of Andrew Davidhazy.)
Photographic Global Notes, Volumes 1 and 2, Tim Mantoani, Dean Collins Productions / Silver Pixel Press, 1994. (In pages 95-98, a camera modified by Brian Lawler is described. His modifications to a Nikon F use the Davidhazy method of a powered rewind to move the film.)
Good Technical Resources
Engineering and Scientific High Speed Photography, William G. Hyzer, Macmillan Company, New York, 1962.
Applied Photographic Optics, Third Edition, Sidney F. Ray, Focal Press, Oxford, 2002.
Cameras From Daguerreotypes to Instant Pictures, Brian Coe, Crown Publishers Inc., 1978.
If you like this story and think others will also, please click on this link to digg it!
(Tom Dahlin is a freelance photographer based in Minneapolis, Minnesota. He covers everything from preps to the pros, and his work has appeared in SI, SI-Kids, USA Today, and ESPN the Magazine. A twenty year stint in the R&D engineering labs of 3M and Honeywell provided him with a unique set of technical skills that he loves to apply to photographic problems. His member page is http://www.sportsshooter.com/tdahlin.)
Dahlin's member page
Digg this story
Related Email Addresses:
Tom Dahlin: firstname.lastname@example.org
Contents copyright 2013, SportsShooter.com. Do not republish without permission.